Training guide
What is HRV?
Heart rate variability, or HRV, is the small variation in time between one heartbeat and the next. A heart beating at 60 beats per minute does not beat exactly once per second. The gaps vary, and HRV puts a number on that variation.
Wearables report HRV mostly during sleep, often as RMSSD in milliseconds. The value differs by device and method, so one number from one night says little. The trend against your own usual range is what researchers look at.
What research says about HRV and training
Studies of endurance athletes have used HRV to monitor how well an athlete is adapting to training. A review focused on elite athletes found mixed results: both higher and lower HRV have been linked to negative adaptation, and in some athletes fitness improved while HRV fell. The authors argue that day to day single readings are hard to interpret and that averaging over several days, and using the right index, give more useful information.
A second review argues that many contradictory findings come from method problems rather than from heart rate measures being useless. It says measures from 5 minute, almost daily resting recordings are likely among the most useful, and that a change should be judged against the measurement error, the smallest important change and the training context.
Why one reading is not the answer
HRV moves with sleep, alcohol, illness, stress, hard training, travel and time of measurement. Wrist and ring sensors measure it in different ways, and brands use different windows and math. The same person can see different numbers on two devices.
So the useful questions are about change: is this week's average inside or outside your usual range, and what happened in training and sleep before it. HRV is not a diagnosis, and Memory.fit does not turn it into a readiness verdict.
What your AI can read in Memory.fit
Progress › Recovery shows HRV and resting heart rate as 7 day averages with your usual range, plus overnight values such as skin or wrist temperature, respiratory rate and blood oxygen when your device sends them. Your AI reads the same data with get_recovery, get_daily_metrics and get_trends, together with sleep and training load.
Memory.fit also lists the scores your device makes, such as Oura Readiness, as the maker's own numbers on the maker's scale. Memory.fit does not score recovery itself.
In your AI
The AI you already use, with your data in Memory.fit.
- ChatGPTChatGPT can schedule tasks, but OpenAI does not say a scheduled task can use a custom connector like Memory.fit. Try it once.Set up ChatGPT
- ClaudeClaude runs scheduled tasks in the cloud, and they use your connectors.Set up Claude
Grok BotGrok Bot keeps working after you close the app, so it can write your weekly review while you train.Set up Grok Bot
MuseMuse checks in when you open a chat. Whether it runs on a schedule with Memory.fit isn't checked yet.Set up Muse
DotsDots are always on, each with its own cloud computer, so your dot can check in on its own.Set up Dots- GeminiGemini checks in when you open a chat. Whether it runs on a schedule with Memory.fit isn't checked yet.Set up Gemini
- GrokGrok checks in when you open a chat. Whether it runs on a schedule with Memory.fit isn't checked yet.Set up Grok
- Perplexity ComputerPerplexity Computer checks in when you open a chat. Whether it runs on a schedule with Memory.fit isn't checked yet.Set up Perplexity Computer
- Claude CodeClaude Code checks in when you open a chat. Whether it runs on a schedule with Memory.fit isn't checked yet.Set up Claude Code
CueYour Cue agents check in from your real training, sleep and recovery.Set up Cue
What to send
Copy one, send it to your AI.
“Show my HRV and resting heart rate from Memory.fit for the last 8 weeks against my usual range, and note any days that fall outside it.”
Tools:
get_recoveryget_trends“Line up my HRV in Memory.fit with my sleep and my training load for the last 4 weeks and describe what changed first.”
Tools:
get_daily_metricsget_sleepget_training_load“Every morning, check what is new in Memory.fit and tell me if my HRV average is outside my usual range, with the training and sleep from the days before.”
Tools:
get_updatesget_recovery“Which device sends my HRV to Memory.fit, and how does the method differ from my other sources?”
Tools:
get_athleteget_source_data
Questions
In short.
There is no single normal. HRV differs a lot between people and between devices, so your own usual range is the useful reference.
Not always. A review of elite endurance athletes found both increases and decreases associated with negative adaptation, and some athletes improved fitness while HRV fell.
Researchers suggest averaging several days rather than reading single values, and judging a change against measurement error and the training context.
No. It shows your HRV, resting heart rate and sleep with your usual range, and your AI can describe the changes. It gives no recovery score.
Devices use different sensors, timing windows and calculations. Compare each device only with itself.
Your next chapter starts with everything you've learned.
On Android now. iPhone is coming: join the waitlist and we'll email you the day Memory.fit is on the App Store.
No spam. One email when it's ready.